• Improve handling of XSD complex type, decimal (https://github.com/databricks/spark-xml/pull/631, https://github.com/databricks/spark-xml/pull/638)
• Restore behavior of ignoreSurroundingSpaces (https://github.com/databricks/spark-xml/pull/637)
• Improve schema inference performance (https://github.com/databricks/spark-xml/pull/660)
• Fix corner case of double/float type inference (https://github.com/databricks/spark-xml/pull/644)

See https://github.com/databricks/spark-xml/milestone/14?closed=1

Note that this is intended to be the final stand-alone release of spark-xml, as it is being incorporated into Apache Spark 4.0.